The Intramural Reseaerch Program at the NIMH is one of only a handful of sites in the world in which neural activity can be compared directly with simultaneously and/or sequentially recorded fMRI signals. We have integrated electrophysiology and imaging in alert, behaving monkeys by collaborating with the Neurophysiology Imaging Facility, a shared imaging facility dedicated to structural and functional brain imaging in nonhuman primates. In previous work, we have explored the relationship between neural signals and fMRI in the primary visual cortex of monkeys. Last year we published a study that for the first time identified specific conditions under which the two signals diverged. While fMRI and neural signals were normally in perfect sync, we showed that they behaved very differently during perceptual suppression, where a stimulus was seen but not perceived. This finding is an important link toward interpreting the results of many human neuroimaging studies that seem to disagree with electrophysiological recordings. In the last few months, we published a study investigating the neural basis of the resting-state fMRI signal. This signal, corresponding to the spontaneous, endogenous fMRI fluctuations that occur in the brain when a subject is not performing any explicit behavior, is studied widely in the human brain by hundreds of laboratories. It is of great interest because the statistical relationship of spontaneous fluctuations measured at different points in the brain carries information about the brains functional processing, a phenomenon termed functional connectivity. We were interested in the neural underpinnings of this phenomenon, and therefore performed simultaneous electrophysiological and fMRI measurements in awake monkeys. Surprisingly, we found that not only is the electrical activity of the cerebral cortex correlated with specific functional circuits, it is also correlated with activity over large swathes of the cortex. This nearly global span of signal fluctuations is an important aspect of brain activity that has been ignored, literally discarded, by the human neuroimaging community as noise. Our findings suggest that it is not noise, but may, in fact, represent that aspect of brain function that accounts for its highest fraction of metabolic consumption. These findings will influence the manner in which the human neuroimaging community considers and treats the global fMRI signal measured during the resting state. In a second, ongoing project, we have measured fMRI signals following focal brain injury. Specifically, we made targeted ablations in the primary visual cortex (V1) of nonhuman primates, and then observed the extent to which fMRI responses in areas receiving input from V1 reemerged after several weeks. We are particularly interested in whether the recovery of the fMRI signal shows the same properties, including the basic recovery time course, as the responses of single neurons measured in the same part of the brain. This approach therefore requires the careful coordination of fMRI, ablation, and electrode implantation. This three-pronged approach is then combined with behavior, asking the animal to tell us when they detect a visual target, to determine how the neural and fMRI signals related to one another, and how they further relate to perception. As a final portion of this project, we have investigated the neural pathways by which information may bypass the primary visual cortex. This was achieved by blocking electrical activity in an intermediate, relay nucleus, and then using fMRI to measure responses. Some of these results, which outline an important pathway mediating V1-independent vision, were recently published.